Related art is disclosed in, for example, Patent Literature 1 through 3 for detecting deterioration in a communication quality at a TCP (Transmission Control Protocol) level as a session control protocol in an IP (Internet Protocol) network.
Patent Literature 1 discloses a method of measuring traffic of TCP by a measuring device provided between a transmission terminal and a reception terminal, calculating a logical throughput of the TCP from a network delay, a packet loss rate or the like measured by the measuring device, comparing the calculated logical throughput and an actual throughput and when the actual throughput is lower than the logical throughput, determining that performance is deteriorated.
Patent Literature 2 discloses a method of determining that communication is deteriorated when the number of TCP retransmission time-outs exceeds a threshold value according to a time and a procedure of a TCP packet.
Patent Literature 3 discloses a method of measuring a reciprocation delay time of a transmission terminal and a reception terminal and determining whether a response delay derives from a server terminal or a network based on three sensing results, a drastic increase in a reciprocation delay time, a slow increase of the same and a decrease in an advertisement window (the amount of use of a reception buffer on reception side) below a value set in advance.
Among related art for determining deterioration by a network repeater are, for example, the techniques disclosed in Patent Literature 4 and 5.
Patent Literature 4 discloses a traffic congestion symptom monitoring system which when traffic information (the amount of use of a communication buffer) in a network exceeds a predetermined threshold value, notifies to that effect.
Patent Literature 5 discloses a device which monitors a dedicated line transmission queue and when the transmission queue overs a predetermined length, determines that a flow of a transmission packet exceeds a dedicated line speed to distribute a part of packets to an ISDN line.
Patent Literature 1: U.S. Pat. No. 3,602,972
Patent Literature 2: Japanese Patent Laying-Open No. 2008-141777
Patent Literature 3: Japanese Patent Laying-Open No. 2006-180002
Patent Literature 4: Japanese Patent Laying-Open No. 2003-249959
Patent Literature 5: Japanese Patent Laying-Open No. H05-48653.
The methods disclosed by Patent Literature 1 and 2 have a problem that deterioration in communication quality cannot be determined in a section where no packet loss occurs. The problem could occur not only in communication using a packet but also in communication using data such as a cell or a frame.
The reason is that they are the determination methods premised on that communication quality deterioration is caused by generation of a packet loss. Even when no packet loss occurs, communication quality might be deteriorated due to a narrow link bandwidth with respect to a data input rate or due to a decrease in a speed-up rate in a session congestion control system under a large delay circumstance.